Lap shear test with silicone adhesive
TBau
PisaMember Posts: 10
Dear community,
For my thesis I am modeling (3D) a lap-shear test for an adhesive-bonded joint. Two metal plates (one steel, one aluminium) - bonded together by a silicone bead - are clamped at the machine at their ends. The lower clamp is fixed, while the upper one moves upward. As during the experiment we reached large displacement, a hyper-elastic Ogden1 material model is set for the silicone, and Large Deflection is switched ON in the analysis setting. For the elements, I chose SOLID185 adding an u-P formulation to avoid distorsion. For the bonded regions, I chose not to merge the geometry on space clam, but to define a comtact region in the modeler. The contact region is "bonded" with MPC formulation.
What I am trying to get is the strain along three path in Z direction (find attached):
- very close to the bonded region between silicone and the steel plate
- very close to the bonded region between silicone and the alum. plate
- in the middle of the silicone bead
Despite the global force-displacement curve really well follows the experimental one, I have some problems when I check the solution: should I check the strain values on the SCS or on the GCS? The SCS gives me negative values in the middle of the silicone bead, while the GCS positive values. Furthermore, both solution disagree with the displacement field: if I try to derive the strain from the displacement in Z direction along the path, the solution is closer to reality.
How can I solve this problem? Should I define another type of CS? Is possibile to have a CS which follows the deformed shape of the path point by point?
Thank u!
Comments
Please define your abbreviations.
Yes, sorry! SCS= solution coordinate system
Please find attached:
- the elastic strain (z direction) in the middle plane of the sample, according to the GCS
- the elastic strain (z direction) in the middle plane of the sample, according to the SCS
- A graph with the elastic strain (z direction) along the centerline, according to both GCS and SCS
ANSYS staff are not permitted to open attachments. Please insert image directly into the post using the Insert Image button.
Above you find the attachements. The first pic was damaged, but it was just the mesh. You can still se a focus on the right.
@zahaa, don't post a new question unrelated to the discussion, create a New Discussion. I have moved your question here.
@TBau, I assume the z-axis is vertical in the images of the mesh above. I still don't know how you defined the Solution Coordinate System.
Yes, it is vertical. I did not define the Solution Coordinate System by myself: it is provided automatically in the output analysis options. As this is a non-linear FE analysis, with Large Deflections + Material Nonlinearities, I think I should not check the strain in the Global Coordinate System - as I do with Linear FEA and Small Displacement, where the strain in z direction is simply e_z=d(u_z)/dz - but in a Coordinate System which follows the deformed shape of the elements step by step. Because with Large Deflection the strain referes to the previous configuration, not to the intial one. What I am looking for, is a definition of a Coordinate System which is able to provide the strain value along a path, simulating the optical fibers that were embedded in the silicone during the test. Do have any tips? I hope I made myself clear.
Not tips? In other words, I am asking in which coordinate system I should check the solution when I want to probe the normal strain in a certain direction in an analysis result performed with Large Deflection ON. And if there´s a way to probe the normal strain along a path in the direction of the path (point by point).